zfs discuss - Jul 2009 - Zfs destroy taking inordinately long time...

I ran a zfs destroy on a 20TB volume on a Thumper running snv_117, and its been
2 hours now with a huge amount of read activity. In the past(2008.06) destroy
came back with in a minutes.
After a couple of hours , activity still looks like:

----------  -----  -----  -----  -----  -----  -----
dataPool    18.8T  21.2T    364      0  2.01M      0
rpool       26.5G   902G      0      0      0      0
----------  -----  -----  -----  -----  -----  -----
dataPool    18.8T  21.2T    342      0  1.95M      0
rpool       26.5G   902G      0      0      0      0
----------  -----  -----  -----  -----  -----  -----
dataPool    18.8T  21.2T    348      0  1.98M      0
rpool       26.5G   902G      0      0      0      0
----------  -----  -----  -----  -----  -----  -----


Is this expected in snv_117? and if not , how to start debugging?

Dan
-- 
This message posted from opensolaris.org

On Mon, Jul 06, 2009 at 09:08:44AM -0700, Daniel Liebster
wrote:
> I ran a zfs destroy on a 20TB volume on a Thumper running snv_117, and its
been 2 hours now with a huge amount of read activity. In the past(2008.06)
destroy came back with in a minutes.
> Is this expected in snv_117? and if not , how to start debugging?
My experience: the more filesystems the longer does it take to
create/destroy/rename a zfs or initially list -r a zfs.

E.g. S10u7 with a pool of 2208 zfs'':
time for create/destroy ~20 seconds/zfs 
time for rename: ~40 seconds/zfs

(Generic_139556-08, X4600M2, 4x DC AMD Opteron 8222, 2993 MHz, 16GB)

Regards,
jel.
-- 
Otto-von-Guericke University     http://www.cs.uni-magdeburg.de/
Department of Computer Science   Geb. 29 R 027, Universitaetsplatz 2
39106 Magdeburg, Germany         Tel: +49 391 67 12768

The destroy process must have hit  point in the FS with a hundred thousand
files. The destroy completed relatively quickly after passing that point.
Please disregard this post.
-- 
This message posted from opensolaris.org